Development and control of molybdenum crystals in a stoneware glaze


Material Information

Development and control of molybdenum crystals in a stoneware glaze
Physical Description:
v, 87 leaves : col. ill., graphs ; 28 cm.
Bassett, Dorothy Rita, 1949-
Publication Date:


Subjects / Keywords:
Glazes   ( lcsh )
Crystal growth   ( lcsh )
Molybdenum   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (M.F.A.)--University of Florida.
Includes bibliographical references (leaf 86).
Statement of Responsibility:
by Dorothy Rita Bassett.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 04379899
System ID:

This item is only available as the following downloads:

Full Text
Dorothy Rita Bassett

I would like to express my appreciation to the members of my thesis committee, Mr. Phillip Ward, Dr. Larry Hench, and Mr. Eugene Grissom for their assistance in the preparation of this manuscript.
I am grateful to the members of the department of Materials Science Engineering, Larry Hench and Paul Johnson, also Ed Ethridge, Craig Baker and Fumio Ouichi, for their interest, help and generous contributions of equipment time.
Very special thanks to Katy King for her interest, assistance, encouragement, enthusiasm, intellectual stimulation and fellowship.
To my husband, Richard, thank you for your constant encouragement and support.

ACKNOWLEDGEMENTS......., ,.............................. ii
ABSTRACT................................................ V
INTRODUCTION........,................................... 1
The Original Glaze Composition.............. 9
Additions to the Glaze...................... 10
Additions of Coloring Oxides to the Glaze... 22
DEVELOPMENT................................. 40
Kiln Atmosphere............................. 4 0
The Firing Cycle............................ 41
Clay Body Composition........................ 45
Preparation for Glaze Flow.................. 45
Choice of Shape for Use with the Glaze...... 47
Fuming...................................... 48
Experimental Instruments.................... 51
Experimental Procedures..................... 55
Experimental Results........................ 57

Discussion of Results..................... 70
CONCLUSION............................................. 75
APPENDIX............................................... 77
BIBLIOGRAPHY........................................... 86
BIOGRAPHICAL SKETCH.................................... 87

Abstract of Thesis Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Fine Arts
Dorothy Rita Bassett
June, 1978
Chairman: Phillip A. Ward Major Department: Fine Art
A stoneware glaze was developed which, upon the addition
of molybdenum trioxide and titanium dioxide, produced an
unusual crystal formation. The compositional elements and
firing conditions necessary to optimize crystal growth were
empirically determined. Glaze defects were analyzed and
corrected, and considerations for using the glaze on ceramic
ware were examined. The crystal phase was characterized
qualitatively and quantitatively.

The production of a macro-crystalline glaze on ceramic ware represents a controlled manipulation of a glass system. Although crystalline glazes have been produced since early in the twentieth century, investigators have consistently noted the difficulty of acquiring controlled effects with these glazes.
Early in the century those investigating the formation of crystals most frequently examined the chemical composition of crystal producing glazes. In 1937 F. H. Norton published an important study which linked temperature and firing conditions to the controlled production of a variety of desired effects in a crystalline glaze.''" All studies concerning these glazes suggest that crystals are delicate and elusive in nature. The chemical composition of the glaze which contains them must be finely balanced, and the firing conditions in which they grow must be properly controlled. The
production of macro-crystals, then, has always presented a challenge to the ceramic artist. The study of such a glaze
F. H. Norton, "The Control of Crystalline Glazes", Journal of the American Ceramic Society, XX, 1937, 217-224.
Macro-crystals are those crystals large enough to be discerned with the naked eye.

system provides the investigator with an unusually detailed look at the chemistry of a glass system and at its behavior during firing.
Most of the crystalline glazes which have been investigated to date are those containing the zinc-silicate crystal. Literature pertaining to the production of these large and often impressive looking crystals is now relatively abundant, and most ceramic artists have the technical capability to produce them. The structure of the zinc-silicate crystal has been examined in detail by Norton and others.
While zinc oxide is the most common crystallizing agent used in a glaze, it has been observed that oxides of other elements such as titanium, magnesium, iron and molybdenum also have a tendency to crystallize. For the purposes of this investigation molybdenum oxide was used as the primary crystallizing agent in the glaze system. The glaze which was developed grew crystals which, though of unknown structure and character, were definitely associated with the presence of molybdenum oxide in the glaze. Until it has been characterized, this crystal shall be referred to as a "moly crystal". The moly crystal has not been extensively studied, and there is little literature referring to the use of molybdenum compounds in a crystalline glaze. The moly crystal is markedly different in character from the more commonly produced zinc-silicate crystal. The unusual shape, subtle color characteristics and elusive character of the moly crystal make it particularly attractive to this author.

The purpose of this investigation has been to develop glaze containing moly crystals which is practical for use by the ceramic artist. This study will attempt to determine the compositional qualities and firing conditions necessary for production of the moly crystal. In addition, the heretofore uncharacterized moly crystal will be identified and structurally and compositionally analyzed in order to provide a better understanding of its crystal habit and behavior.

X-Ray diffraction studies show that glass consists of tetrahedral molecules in which one silicon atom is linked to four oxygen atoms. These basic units join to form a three-dimensional network of randomly spaced molecules. The crystal, in contrast to amorphously structured glass, is a solid composed of molecules arranged in a regularly repeating three-dimensional pattern.
Tammann's work on crystalline glazes'*" indicates that there
are two steps in the development of macro-crystals in a glaze;
the formation of nuclei, resulting from the introduction of
particular oxides into the glassy matrix, and the growth of
these nuclei. There is some disagreement as to the exact
nature of nuclei. While some hypothesize that nuclei are
actually tiny crystals, others characterize them as stable
complexes of ions and molecules which are capable of further 2
growth. The formation of nuclei can only occur when the glassy matrix is heated to the point where thermal agitation will stretch and break the electronic bonds which bind the
F. H. Norton, "Control of Crystalline Glazes", 218.
L. B. Hennessey, Factors Affecting the Growth of Zinc-Silicate Crystals in a Glaze, Masters Thesis, University of Puget Sound, 1974, 4."

molecular structure of the glass. In this fluid state,
silicon dioxide and the stabilizing oxides which normally
fill the spaces within the glassy network can move about
freely, allowing the attachment of those atoms necessary
for crystal formation and the migration of unwanted atoms
away from the crystal. On the basis of chance, in proper
thermal conditions, atoms will group to form unit cells,
creating nuclei around which crystals can grow. Thermal
conditions must be precisely controlled for the development
of these unit cells. If the temperature is too high thermal
agitation will dissolve the nuclei as soon as they form; if
too low, the atomic bonds may be too rigid to permit the
atomic rearrangement necessary to produce perfect unit cells.
Nuclei must grow in order to produce macro-crystals.
Norton suggests that for some crystals, growth can take place
in the same temperature range in which the nuclei are formed.
Other crystals require one temperature for nucleation and
a higher one for growth. While most investigators recommend
a soaking period or slow cool within the growth range of a
crystalline glaze, Sanders suggests that some glazes will
form macro-crystals with a normal cooling cycle.
F. H. Norton, "The Control of Crystalline Glazes", 218.
H. Sanders, Glazes for Special Effects, New York, 1974, 43.

MOLYBDENUM OXIDE AND ITS USE IN CERAMIC PROCESSES Molybdenum is a silver-blue metal found most commonly in the molybdenite mineral but also occasionally derived as a by-product of copper ore. Molybdenum tri-oxide, MoO^, the form of the element most often used in ceramic processes, is a white powder which consists of rhombic crystals. MoO^ may be reduced to a lower form of the oxide, M0O2/ caus^-n9 it to take on a characteristic blue or green tint. When sublimed, molybdic oxide forms rhombic tablets and thin needles which exhibit the interference patterns of thin films which are highly light dispersing and double refracting."'" When heated, molybdic oxide begins to pass from a solid to a gaseous state at 1292 0 F (ca. 700C), melting at 1463F (ca. 798C), and boiling at 2111F (ca. 1154C). Traces of basic oxide material, i.e., the conspicuously fluxing agents of the RO, R2O group, are thought to reduce the vapor pressure
of molybdic oxide, providing an explanation for the fact that
Mo07 remains in a glaze when fired above its boiling point.
M. J. Alstrom, The Development of Molybdic Oxide Crystals and Their Use in Decoration on Porcelain, Masters Thesis, San Jose, California, 1960, 4.
M. J. Alstrom, Development of Molybdic Oxide Crystals, 5.

Molybdic oxide is soluble in water and has traditionally
been used in glasses and enamels as an agent for melting and
reduction of surface tension. It has also been used as an
opacifier in enamels and metals and as a coloring agent in
glass and glazes, imparting yellowish-blue tones at tempera-
tures as high as Orton standard pyrometric cone 12.
To date there has been scant investigation into the
possibility of growing moly crystals in a ceramic glaze.
Molybdenum is mentioned in the literature, along with zinc,
titanium, magnesium and iron, as an element which has a
pronounced tendency to crystallize. Parmelee suggests adding small amounts of molybdenum to promote the formation of other types of crystals in a glaze.
Sanders reports success in growing moly crystals in glazes ranging in maturing temperature from cone 3 (2134F, 1168C) to cone 10 (2381F, 1305C).6 They appear to grow in glazes which contain members of the RO, R2O group which are high in atomic weight such as lead and barium. This
Pyrometric cones are small triangular cones, 1 1/8 by 2 5/8 inches in height, made of ceramic materials which are compounded to melt at a specific combination of temperatures and time. G. Nelson, Ceramics, New York, 1971, 377.
L. B. Hennessey, Factors Affecting Growth of Crystals, 6.
C. W. Parmelee, Ceramic Glazes, Boston, 1973, 545.
H. Sanders, Glazes for Special Effects, 43.

finding contradicts Stull's observation that crystal form-
ation is hindered by the presence of these heavy elements.
Kahn reports the formation of six distinct crystal shapes in a stoneware glaze containing molybdenum in various forms, but there is no certainty that all of these formations
are related to the presence of molybdenum.
While Alstrom and Kahn have successfully grown and described the irridescent, rhombic crystal which is associated with molybdenum, the properties and characteristic behavior of this crystal remain to be elucidated.
C. W. Parmelee, Ceramic Glazes, 544.
A. Kahn, Molybdenum as a Crystal Producing Agent in a Stoneware Glaze, Masters Thesis, Texas Woman's University, 1970.
Kahn reports that at least one of the crystals appeared without the addition of molybdenum. Furthermore, the descriptions of other of the crystal formations might also describe forms of titanium crystals. No analysis has been done to determine their structure or composition.

The Original Glaze Composition Several glazes containing molybdenum trioxide were tested with a series of firing schedules. While some produced small crystals of various types, only one showed the rhombic crystal formation associated with molybdenum. This glaze, denoted as CF1, was a variation of a glaze mentioned by Herbert Sanders.^ Empirical Formula of Glaze CF1
.2 B203 .47 Si02 2.05
.04 Al23 ,18
Na20 CaO BaO ZnO
03 23
Batch Formula of Glaze CF1 Custer Feldspar. Gerstley Borate Whiting
Barium Carbonate Zinc Oxide
Edgar Plastic Kaolin
32.49 23.85 6. 86 1.79 6.76 3.69 2.45 22.2
H. Sanders, Glazes for Special Effects, 41,

Additions to the Glaze
Varying additions of molybdenum and other crystallizing agents were made to Glaze CFl to evaluate their effect on the crystal's formed. These additions are listed in Table 1. x
The glazes were applied by brushing and dipping on
porcelain test tiles. The tiles were fired in an oxidation
atmosphere in a Globar silicon carbide element kiln.
Although the firing was intended to go to cone 9 (2336F,
1280 C), a mistake in firing procedure caused the kiln to
overfire at a rapid rate to cone 11 (2300 F, 1315 C) The
firing was completed, however, and the kiln was slowly cooled
to 1575F in six hours, then allowed to cool normally.
The fired tiles exhibited a very glassy, excessively fluid glaze. Many tiles showed a profusion of blisters and pinholes. It was suspected that the excessive running was caused by overfiring the glaze, and that the pinholes resulted from too rapid a heat rise in the kiln.
Crystals were evident on tiles glazed with variations B, C, D, E, F, H, J, L, M, N and O in Table 1. Glazes B, C, J, L and O varied from clear to white to bluish white, and the crystals did not differ in color from the background glaze. The crystals were generally square in shape and quite small, growing to a maximum of only 1/8 inch in diameter. The crystals seemed to resemble tiny versions of the rhombic
See Appendix, Firing Schedule 1.

Table 1. Additions to Glaze CF1.
A. 4% Mo03
B. 4% 03 + 1% Wo03
C. 4% 03 + 1% Wo03 + 8% Tio2
D. 4% 3 + 2% Wo03
E. 4% 03 + 2% woo3 + 4% Tio2
F. 1% Wo03
G. 2% woo3
H. 2% woo3 + 8% Tio2
I. 1% Wo03 + 8% Tio2
J. 4% Mo03 + 4% Tio2
. 1% 3 + 1% woo3
L. 4% 3 + 6% Tio2
M. 8% 03 + 2% woo3 + 6% Tio2
N. 2% 03 + 6% Tio2
0. 4% 03 + 8% Tio2

crystal associated with the presence of molybdenum as
3 4
described by Sanders and Kahn. Additions of molybdenum
alone did not produce crystal growth. The presence of 5
tungsten and titanium appeared to promote crystal formation, and subsequent repetition of this test series suggested that additions of 2 to 8 percent titanium produced particularly good results. As the amount of titanium increased, the glaze became increasingly mottled and blue in color.
Glaze variations D, E, F, H and M exhibited tiny crystallike white specks which were rough in texture. These variations were fired several times on varying firing schedules with no success in producing crystal growth. No further experimentation was carried out with glaze variations A, B, D, E, F, G, or H listed in Table 1. Second Firing Cycle
Glazes C, J, L and 0 were fired in a silicon carbide element electric kiln, held at peak temperature for thirty minutes and cooled at a rate of approximately 50F/thirty minutes to 1550F, then cooled normally to room temperature. The glazes were applied to porcelain test tiles as before, with the exception of glaze O which was applied to a bowl form.
H. Sanders, Glazes for Special Effects, 41
A. Kahn, Molybdenum as a Crystal Producing Agent, 22.
Tungsten is also known as Wolfram, Wo.
See Appendix, Firing Schedule 2.

Small square crystals were evident on all tiles. Variation C exhibited an unpleasant, rough surface. Glaze variation 0, applied to a bowl form, showed crystals of up to 3/4 of an inch in diameter, clustered along the lower sides of the form and pooled at the bottom of the bowl.
The crystals appeared in two distinct forms, with a number of intermediate variations. Some crystals were square in shape, varying from 1/8 inch to approximately 1/2 inch in diameter. These crystals were highly lustrous and light dispersing, appearing to be made up of concentric bands of rainbow colors.
As the crystal formations exceeded 1/2 inch in diameter, the four corners of the crystal began to elongate. The largest of the crystals might be described as having a pinwheel or flower-like shape with four distinct arms radiating from a central point. One of the arms frequently grew to be longer than the other three. Each of the four arms was bisected longitudinally, and each half of the arm reflected light at a different angle. A distinct border surrounded the entire crystal (Figures 1-4).
The crystals had no distinct color of their own but were translucent, showing the color of the background glaze through what appeared to be a thin, irridescent film. The larger, pinwheel configuration crystals extended above the surface of the glaze and were tactily different from the glassy phase of the glaze. The crystals were relatively soft and could be scratched and marred by a metal instrument.

Figure 1. Moly Crystals.
Largest crystal 1" diameter.

Figure 2. Moly Crystals.
Largest crystal 1" diameter.


Figure 3. Moly Crystals.
Largest crystal 1" diameter,

Figure 4. Moly Crystals.
Largest crystal 1" diameter.


The largest of the crystals sometimes lost the quality of irridescence.
While the crystals which formed in this glaze were most satisfactory, the glaze again exhibited a profusion of blisters and pinholes and was excessively fluid.
Additions of Coloring Oxides
Kransr, in studying coloring oxides in a zinc-silicate crystalline glaze, found that of the seven most commonly used oxides, some were absorbed by the crystal more readily than others, resulting in a color contrast between crystal and glaze magma.
To test the effects of coloring oxides on glaze CF1,
varying amounts of cobalt, iron, manganese, copper, chrome,
uranium and nickel were added to the glaze. Combinations
of these oxides were also tested, as were several Mason
ceramic stains. The specific additions and results of this
series of tests is shown in Table 2.
In no case did the crystal absorb color in a manner significantly different from the background glaze. The
background color showed through the translucent crystal and
mixed with the crystal's irridescent rainbow colors. The
crystal remained subtle and elusive and in some cases was
rtot readily obvious on the glaze surface.
Of the seven coloring oxides tested, manganese produced
the least desirable colors whether used alone or in
H. M. Kraner, "Colors in a Zinc Silicate Glaze", Journal of the American Ceramics Society, 1924, 7:868.

Table 2. Additions of Colorants to Glze CP1.
Colorant % Color Formation
. 1% CoCo3 Pale Blue yes
. 5% CoCo3 Medium Blue yes
.5% CuCo3 Pale Green yes
3% CuCo3 Pale Green yes
2% Ni02 Mottled Green no
2% Fe203 Golden-Orange yes
.5% Cr02 Mottled Greenish-Brown no
io% u2o3 Yellow yes
2% Mg02 Mottled Brown yes
1% NiO +3% CuCo3 Irridescent Green yes
.5 Mg02+2% Fe203 Mottled Brown no
.5% CrO +.5% CuCo3 Mottled Green no
.5 % CoCo3 +.3 % CuCo3 Mottled Blue-Green no
.5% CuCo3+.5% Cr02 Mottled Blue-Green no
*7% Crimson Stain #16 Pale Pink no
*7% Robins Egg Blue Turquoise no
Stain #K-75
*5% Wedgewood Blue Intense Blue no
Stain #10
*7% Black Stain Mottled Black-Blue yes
*3% Pansy Purple Intense Blue no
Stain #85
*7% Deep Salmon Salmon no
Stain #331

Table 2 continued
Colorant %
Crystal Formation
*7% Persimmon Stain #K-1403
?Peacock Blue Stain #1296
*Deep Brown Stain #6151
*Spice Brown Stain #1533
*Walnut Brown Stain #308
*Camel Brown Stain #1532
Intense Blue Dark Brown Dark Brown Dark Brown
Medium Brown
Stains from Mason Color and Chemical Co., East Liverpool, Ohio.

combination with other colorants. Nickel used by itself produced uninteresting results, but when 1 percent nickel was combined with 1 to 3 percent copper, an attractive grey-green with a hint of irridescense resulted. Uranium in the range of 8 to 10 percent produced a yellow which seemed to enhance the color characteristics of the crystal. An addition of 2 percent iron gave the glaze a bronze-gold color with some mottling and a hint of irridescence. Cobalt proved to be such a powerful colorant that additions of .5 percent or more produced a glaze so strong in color that it overpowered the subtle colors in the crystal. Copper produced a pleasant green in amounts of .5 percent to 3 percent.
Kraner observed that additions of as little as .01 equivalents of chromium oxide hindered or prevented crystal
growth in the zinc-silicate glaze he tested. Chromium exerted the same influence on glaze CFM, usually preventing crystal growth entirely.
Combinations of two or more coloring oxides in glaze CFl tended to produce colors which were quite mottled. Though sometimes attractive in themselves, these mottled colors obscured the delicate crystal and were therefore unsatisfactory.
The rather high titanium dioxide content of glaze CFl contributed significantly to the mottled effect. Reducing
H. Kraner, "Colors in a Zinc Silicate Glaze", 877.

the titanium dioxide from 8 percent to 6 percent and 4 percent improved this condition without adversely affecting crystal growth. It was further observed that substituting the mineral rutile, an impure, often iron-bearing form of titanium dioxide, produced slightly warmer colors. The glaze was therefore permanently revised to contain an addition of 4 percent rutile.
The addition of a series of Mason stains to the glaze generally proved to be disappointing. Chrome-tin stains such as Crimson 161, Deep Salmon 331 and Persimmon K-1403 prevented crystal formation entirely. Small, densely packed crystals did grow in chrome-tin stain Peach 211. Those stains containing cobalt, such as Wedgewood Blue 10 and Pansy Purple 85, even when used in very small amounts, produced strong colors which overpowered the crystals. Peacock Blue 1296, probably because it contains chrome, also retarded crystal formation. A series of darker stains was tried, including Deep Brown 6151, Spice Brown 1533, Walnut Brown 308, and Camel Brown 1532. The resulting colors were dark and dominant, and the chrome content of these stains again retarded crystalization. Chrome-free Black Stain 616 produced an attractive, broken blue-black which set off the crystals nicely. Tin-Vanadium Yellow Stain 301 produced little coloration even in relatively large amounts, while
Available from Mason Color and Chemical Works, Inc., East Liverpool, Ohio.

Chartreuse Green Stain 1386 produced an attractive yellow but seemed to retard crystal growth.
It was concluded that small additions of single colorants generally produced the most attractive, least mottled colors which worked to enhance the crystal rather than to overpower it. Crystals grew with greatest regularity in those glazes containing cobalt and copper, to a lesser extent in glazes colored with uranium and iron, and least successfully in glazes colored with chromium oxide.

Blistering and Pinholing
A close examination of the blisters and pinholes found on the bowl fired on schedule 2 with glaze CFl, variation 0, revealed several types of blisters. While some were very small with smooth edges, what are commonly known as pinholes, others appeared as large, sharp edged eruptions which revealed the clay body beneath the glaze. The presence of these varied formations indicate that the blisters formed, erupted, healed and reformed throughout the firing process. They were thus caught and frozen at different stages of formation when the glaze solidified on cooling. Glaze Application
Visual examination of the characteristics of the blisters seemed to indicate that the fault did not lie in application. However, to eliminate this possibility, a series of small bowls and bottles were glazed with glaze CFl, variations J, L and 0 by dipping, pouring, brushing, and spraying. The pieces were fired to cone 9 on schedule 2 in the silicon carbide element kiln.
All fired pieces exhibited blisters and pinholes. The Clay Body
Parmelee states that blisters and pinholes may arise when glaze is applied to raw or very porous ware due to the

adsorption of glaze constituents by the body or the adsorption of soluble salts out of the body into the glaze.^ By bisque firing all test ware to cone 8: (1751F, 955C) or above, these possible causes were eliminated from consideration.
Ball Milling
Glaze CFl with additions of 4 percent molybdenum and
8 percent titanium was ball milled for a short period, two
hours, in order to reduce the particle size of the glaze
constituents and to produce a homogenous glaze slip.
Schramm and Sherwood have found that grinding is an important
factor in determining the properties of a glaze slip and of
the matured, fired glaze.
The ground glaze, fired to cone 9, produced a smooth, fluid glaze which was free from pinholes and blisters. However, crystals did not grow in the glaze.
In another trial, the base glaze was milled for two hours, and the crystallizing agents molybdenum and titanium were added to the milled batch. Again, firing to cone 9 produced a glaze which was smooth and clear, free from pinholes, but which did not support crystal growth. Reducing the milling time to one hour produced similar results.
C. W. Parmelee, Ceramic Glazes, 580.
C. W. Parmelee, Ceramic Glazes, 110.

It was hypothesized that milling, while reducing particle size and homogenizing the glaze, might also decrease the maturation point of the glaze. The smooth, crystal free surface and runny nature of the glaze might then be the result of overfiring. Accordingly, the glaze was fired to cone 8 (2305F, 1263C) on schedule 2. The resulting glaze was semi-matt in surface and contained areas of rough textured white material which might have represented tiny, clustered crystal growth. Although the glaze was free from pinholes, the crystal growth and surface of the glaze were unsatisfactory.
Adjustments in Glaze Chemistry
Zinc and Titanium. Rhodes states that pinholes may
result from excessive quantities of zinc or titanium in the 3
glaze. In order to determine whether this might be the source of the problem in glaze G'Fl, the following variations were formulated:
1. Reduction of Zn02 by 5 percent
2. Reduction of Zn02 by 10 percent
3. Reduction of Ti02 by 5 percent
4. Reduction of Ti02 by 10 percent
5. Reduction of Zn02 by 5 percent and Ti02 by 5 percent. These tests, along with a version of glaze CFl, variation 0, containing calcined zinc oxide, and a control pot
D. Rhodes, Clay and Glazes for the Potter, Raduor, P. 1974, 247.

glazed with CFl, were fired in the silicon carbide element
kiln to cone 9 on firing schedule 2. The results showed
no general improvement. All pieces exhibited a profusion of
blisters and pinholes. The adjustments in the quantities of
zinc and titanium also detrimentally influenced crystal growth
and the surface quality of the glaze. After repeating these
tests in a second firing and obtaining similar results, this
approach to correcting the problem was abandoned.
Boron content. Parmelee suggests that while a small
boron content helps heal pinholes, an excess of approximately
10 percent boron may contribute to blistering and pinholing.
Raw glaze CFl contained boron in the form of Gerstley borate, a form of calcium borate, which comprised 17 percent of the batch weight and contained .48 molecular equivalents of boron. The glaze was revised to contain .2 35 molecular equivalents of boron, while retaining the original level of calcium, and fired to cone 9 on schedule 2 in the silicon carbide element kiln. The results indicated that the reduced boron level did not adversely affect crystal growth. In fact, the crystals from this firing appeared to be particularly lustrous and colorful. The pinholes and blisters were still prevalent, however.
When the boron level was further reduced to .18 molecular equivalents, the quality of the glaze surface began to deteriorate. The optimum boron level was thus established at .235
C. W. Parmelee, Ceramic Glazes, 56.

molecular equivalents, and the glaze batch was permanently revised in accordance with this new formula. This new version of the glaze was denoted as CFM.
Substituting commercial frits. Because the introduction of boron in the form of Gerstley borate was still suspected as a possible contributing factor to the pinholing problem, the glaze was recalculated using two commercially available frits. The frits chosen for use were Ferro Frit 3124 and Ferro Frit 3134. In each calculation the boron requirement was satisfied first. This necessitated the combination of the K20 and Na20 requirements of the glaze into an overall KNaO requirement of .24 molecular equivalents.
Both fritted versions of glaze CFM were fired to cone 9 on schedule 2. Neither proved to be satisfactory. The glaze containing Frit 3134 showed an unpleasant, rough surface which suggested that it was immature. The version formulated with Frit 3124 crazed badly and failed to produce crystals. These experiments seemed to indicate that the chemistry of glaze CFM is delicate and may be easily disturbed. Apparently the levels of K20 and Na20 must be precisely maintained, or, perhaps the introduction of some other glaze ingredient in the frit proved unsatisfactory. Deflocculating the Glaze Slip
It was hypothesized that the pinholing might have been the result of the migration of soluble salts contained in the raw glaze batch across the glaze surface as it dried. These soluble salts were carried by the water mixed in the

slip and deposited on the surface of the glaze as the water evaporated, resulting in areas of glaze which were not of average chemical composition. These areas might react abnormally in the firing and produce blisters. Therefore, in order to decrease the amount of water required to make the
glaze slip fluid, a deflocculant was added to the glaze.
The deflocculant chosen for use was Darvan C, a sodium-free dispersing agent.
When mixed, clay and water normally form a flocculated system, that is, a system in which particles of clay cling together rather than float separately because of the attraction of electrical charges. In most cases, nearly equal volumes of water and dry mixture are required to cause the clumps within the slip to flow freely. In order to decrease the water requirement, the electrical charges which attract the clay particles to one another must be overcome.
The overall electrical charge on a clay platelet is usually slightly negative while water is charged positively with Ca ions. When a polyelectrolyte such as Darvan C is added to the clay-water system, the net positive charge is overcome, and the negatively charged clay platelets repel one another. In the resulting deflocculated system, large amounts of water are trapped between the separated clay particles. When permitted to dry, the system settles into a
Manufactured by R. T. Vanderbilt Co., New York, N.Y.

dense "card-pack" arrangement of ions. Fetteroff found that a deflocculated system acts as "a protective colloid...which, by adsorption on the surfaces of the particles of solids, prevents the solution of soluble salts in water".
A deflocculated system has a higher solid content than a flocculated one, and the drying shrinkage of a deflocculated slip will be lower. A thicker glaze application is therefore possible with a deflocculated slip.
The usual amount of deflocculant added to a slip is approximately 1/3 to 1 percent of the dry weight of the batch. Larger additions usually produce no further effect and may in fact reverse the conditions and bring about a flocculating effect.
Experimental amounts of Darvan C were added to water, and the solution was then mixed with the dry batch of glaze CFM. It was found that Darvan C effectively deflocculated the system at concentrations varying from .25 percent to 1.5 percent.
Since the most effective deflocculating action occurred with 1.5 percent Darvan C water solution, this solution was mixed with glaze CFM and fired on schedule 2 in the silicon carbide element kiln. The results indicated that the addition of a deflocculant was productive. Blistering and pinholing were definitely reduced although not entirely eliminated. The deflocculated glaze produced crystals in
C. W. Parmelee, Ceramic Glazes, 131.

greater numbers and in areas other than the bottom of the test bowl where the glaze usually pooled. It was concluded that this was the result of the extra thickness of the glaze coat which the addition of the deflocculant allowed. Further examination of the accumulated test pieces confirmed the observation that the crystal grew only where the glaze was quite thick. Thus the crystals tended not to appear on the lips and shoulders of objects where the glaze naturally thinned during firing. Instead they collected in the bottoms of bowls or in a thick roll at the foot of a piece. The results of the Darvan C additions gave encouragement to the idea that it might be possible to achieve a glaze coat thick enough that crystals would grow on the vertical surfaces of a pot, an objective which those who have worked with crystalline glazes have continually sought to achieve. Effects of Firing Procedures
In a further effort to eliminate pinholes in the glaze, new firing procedures were initiated. Pinholes sometimes result from the incomplete expulsion of gases which are given off by the glaze during firing. As the gases bubble out of the molten glaze, they leave craters which should smooth over as the glaze matures. If the kiln is turned off while these gases are still being given off, the bubbles will be caught in the solidifying glaze in the process of formation or healing, and will show up as blisters and pinholes. A slower firing time might therefore allow more time for gas to escape and the blisters to heal.

Two batches of glaze CFM were mixed and applied to test
ware, one batch was mixed with water and the other contained
a 1,5 percent Darvan C solution. The pots were fired in the
silicon carbide element kiln on schedule 3. The results were promising. In most cases, the pinholing on these pieces was less severe than on those fired on schedule 2. The group glazed with the deflocculated slip definitely showed fewer blisters than the other group. From this time on in the study, all pots were fired on schedule 3, and a 1.5 percent solution addition of Darvan C was added to each batch of glaze. Kiln Effects
The deflocculated glaze was fired to cone 9 on schedule 3 in an Olympic electric kiln and in an L+L electric kiln, both containing Kanthal heating elements. Both firings produced pots which were totally free from pinholes. After repeating these results it was concluded that the silicon carbide element kiln somehow contributed to the pinholing problem. A possible explanation for this effect might be the fact that the bag walls had been removed from the Globar kiln. The pieces were thus exposed to direct radiant heat from the silicon carbide elements. While most glazes are not noticeably effected by this, it appears that glaze CFM is particular ly sensitive to firing conditions.
See Appendix, Firing Schedule 3.

Excessive Glaze Flow
Crystalline glazes are normally quite fluid, and the potter must take precautions to avoid having the excess glaze seal the pot to the kiln shelf, thereby potentially ruining both pot and shelf. In addition, while it is usually relatively easy to collect crystals on the bottom of a bowl, it is difficult to control the fluidity of the glaze enough to allow crystals to grow on a vertical surface.
Norton notes that the only way to produce a limited number of nuclei which can subsequently grow to macro-crystals is to heat the glaze above the growing range of the crystal until all but a few nuclei are dispersed by thermal agitation,
then to drop the temperature and hold it in the growing
range of the crystal. At peak temperature, when nuclei are being dispersed and reduced in number, the glaze must be quite fluid. One of the conditions required for crystal growth in glaze CFM was a pronounced thickness of glaze coat which further contributed to the tendency of this crystalline glaze to run badly. Alumina Level
The problem was first approached by an examination of the alumina content of the glaze. The presence of alumina in a glaze adds viscosity to the melted glaze and makes it less apt to run off vertical surfaces. However, Parmelee
F. Norton, "Control of Crystalline Glazes", 221.

points out that in the case of crystalline glazes, the presence of alumina can be disadvantageous, as it may control
the rate of crystal formation and crystal size as well as
the flow of the glaze coat. Rhodes states that crystalline glazes are the only glazes commonly formulated without alumina. Snair suggests that the alumina content of a crystalline glaze should not exceed .12 molecular equivalents.
Although glaze CFM contained .18 molecular equivalents of alumina, the level was experimentally raised to .2 05 in order to create a 1:10 alumina:silica ratio. The glaze sample was applied to test ware and fired on schedule 3 to cone 9. The fired glaze showed little improvement in running, and fewer crystals developed than on a control pot glazed with CFM containing .18 molecular equivalents of alumina. This approach to solving the problem was abandoned. Adjustments in Firing Cycle
More success in reducing the flow of the glaze resulted from lowering the peak firing temperature. It was discovered that glaze CFM is mature at cone 8 (2305F, 1263C). Although the glaze was somewhat more opaque and slightly different in color when fired to this temperature, running was substantially
C. W. Parmelee, Ceramic Glazes, 192.
D. Rhodes, Clay and Glazes, 88.
D. Snair, "Making and Firing Crystalline Glazes", Ceramics Monthly, December 1975, 24.

reduced. Thus, if the glaze application was thinned as it approaches the bottom of a piece, ware could be successfully fired without excessive glaze flow.

Kiln Atmosphere
In order to observe the effects of kiln atmosphere on crystal formation, test ware coated with glaze CFM was fired in a salt kiln, and in both oxidation and reduction conditions. Sodium Atmosphere
Glaze CFM was fired to cone 9 in the sodium atmosphere of a salt kiln. During the final hour of salting, the burners of the kiln were turned back in order to produce a gradual cooling effect in the kiln. When salting was completed, the kiln was turned off and allowed to cool normally. The fired piece showed an extremely glassy, intensely blue glaze. There were no crystals apparent anywhere on the piece.
The sodium atmosphere of the kiln clearly acted as a strong fluxing agent, causing the glaze to mature at a lower temperature than usual. It is doubtful, however, that this fluxing action alone accounts for the lack of crystal development. While most crystals might have flowed off the exterior of the overfired pot, any which had formed on the interior should have collected in the bottom of the test bowl. The cooling of the kiln could not be closely controlled, and it is possible that the temperature was not maintained in the growing range of this crystal for a long enough period to produce macro-crystals. The high sodium atmosphere of the

kiln may also have interfered in some way with the crystallization process. Reduction Atmosphere
Glaze CFM was fired in a reduction atmosphere in a 16 cubic foot gas kiln. When fired to cone 9 the glaze became extremely mottled, breaking in color from yellow to blue in a manner typically associated with titanium. No macro-crystals were observed. The difficulties of maintaining a slow, controlled cooling cycle in a large gas kiln may account for the absence of crystal formation. Until temperature conditions in a reduction atmosphere can be closely regulated, it is not possible to observe in isolation the effects which reduction might have on the formation of moly crystals.
Oxidation Atmosphere
When fired on a proper schedule, moly crystals grew consistently in an oxidation atmosphere. The glaze was mature at cone 8, and produced crystals when fired as high as cone 11, but as the firing temperature increased and the glaze became more fluid, the crystals tended to flow off vertical surfaces. At cone 8 colors in the glaze tended to be flat and opaque. As the temperature increased, the color became slightly more transparent.
The Firing Cycle
While Sanders indicates that crystals will sometimes grow in a glaze which is heated and cooled in a normal

manner,^ most crystalline glazes require a controlled cooling
cycle. Norton first related the regulation of the cooling
cycle to the production of specific crystal shapes in a
zinc-silicate glaze. Experimentation with glaze CFM revealed that optimum effects can be achieved only through careful regulation of both the heating and cooling cycle.
Most authors reporting on the production of zinc-silicate crystals do not suggest a special relationship between the rate of the heating cycle and the quality of the fired glaze. Kraner, however, found that crystallization in the zinc-silicate glaze he studied seemed to be more dependent on a somewhat rapid temperature increase than on close control of the cooling. Kraner suggested that a quick heat rise
allowed all the glaze ingredients to be carried into solution
before the mass had an opportunity to flow off the piece. Comparison of glaze CFM fired on schedules 1, 2 and 3 revealed that crystallization in this glaze was not greatly affected by the rate of temperature rise. However, it has been established that this glaze requires a slow heat rise in order to avoid pinholing. It is probable that such a
H. Sanders, Glazes for Special Effects, 43.
F. H. Norton, "Control of Crystalline Glazes", 222.
H. Kraner, "Colors in a Zinc-Silicate Glaze", 869-870.

slow firing contributed to the fluidity of the glaze- It was therefore necessary to determine the heating cycle which would both minimize fluidity and still prevent pin-holing and blistering.
Experimentation to this point had suggested that a slow cooling cycle might be important to the formation of large, well shaped crystals in glaze CFM. This observation was confirmed by experimentally firing the glaze to maturity and then allowing it to cool normally. While the fired glaze
contained a few small, pale circular formations which resem-
bled what Sanders describes as "ghosts", it contained no macro-crystals.
A series of experimental firing cycles was initiated in an Olympic electric kiln to determine the optimum heating and cooling rates for glaze CFM. The specific schedules and recorded results appear in Appendix I, schedules 4-8.
This series of firings revealed several things. Schedule 4 represents the maximum rate of heat-use attempted. The resulting glaze surface was free from pinholes, and this rate was thereafter adopted for all firings.
Schedules 5-8 represent variations in the cooling cycle which determined the limits of the growing range for this particular crystal. While crystals did grow when the cooling cycle was stalled for several hours as low as
H. Sanders, Glazes for Special Effects, 18.

1850F (ca. 1000C), the crystals which formed were ill-defined and imperfectly shaped, indicating the proximity of the lower limits of the growing range. The glaze was quite viscous when fired on this schedule.
When taken to maturity at cone 8, dropped to 2100F (1150C) and held for several hours, the glaze produced masses of tiny crystals. These white, speck-like crystals clustered together so closely in some areas that the glaze surface became raised and rough in texture. The glaze was also particularly fluid when fired on this schedule, probably because at 2100F the glaze was still quite active.
The best results were obtained when the glaze was heated to maturity (cone 8) then quickly dropped below 2000F (ca. 1100C) and held between 1900F (ca. 1040C) and 2000F as in Firing Schedule 7. Unlike the zinc-silicate crystal, the moly crystal does not continue to expand in size as it is held longer in the growing range. The moly crystal rarely exceeds one inch in diameter under any firing conditions. The maximum soaking time necessary to produce crystals of this size seems to be four to five hours. Smaller, square irridescent crystals will grow in as short a soaking period as two hours.

Clay Body Composition Clay body composition can have an effect on the formation of crystals in a glaze. According to Sanders, a single glaze can be matt or shiny, producing crystals in one case but not in another, depending on the clay body to which it is applied.''"
While crystalline glazes will work on both porcelain and stoneware, the body should be quite smooth. If the clay contains sand or grog, these rough particles will show through the crystal formations, disrupting the visual effect. The subtle nature of the moly crystals makes the use of a smooth, flawless body especially desirable.
The color of a clay body may somewhat alter the effect of a crystalline glaze. Since most crystalline glazes are glossy and translucent, the use of a white porcelain body may result in brighter, more vibrant colors. Darker stoneware bodies proved to be particularly detrimental to the moly crystal as it is easily overpowered by strong or dark colors.
Preparation for Glaze Flow
While glaze CFM may be made more viscous than many crystalline glazes by careful control of the firing procedure,
H. Sanders, Glazes for Special Effects, 24.

4 6
some precautions should be taken to avoid damaging ware and kiln furniture should the glaze run.
Glaze should be applied in a way which anticipates the possibility of glaze flow. Thus the glaze coat should be thinned as it approaches the foot of a piece. This may be easily accomplished if the glaze is applied by spraying. If the glaze is applied by dipping or pouring, the dry glaze may be rubbed or lightly sanded off the area near the foot of the pot. A good margin of unglazed area above the foot should always be maintained.
The design of a particular piece will naturally affect
the tendency of a glaze to run. Glaze is most likely to
flow from a vertical surface, and some method must be employed
to prevent the glaze from cementing the pot to the kiln shelf.
While many methods have been devised, one of the most success-
ful ones was suggested by David Snair. A small base which matches the diameter of the foot of a piece may be thrown or fashioned by hand out of the same clay body as the pot. Pot and base are bisqued together. After glaze has been applied to the pot, the base is attached with a thick mixture of alumina hydrate, white glue, and a small amount of water. This mixture tightly seals the base to the pot. Glaze will then flow off the foot of the pot and down the base without seeping under the foot. A substantial dusting of kaolin on
D. Snair, "Making and Firing Crystalline Glazes", 23-24.

the kiln shelf will absorb any glaze that flows from the base. After firing, a chisel will easily separate base from pot, and any rough edges of glaze around the foot of the piece may be ground away without difficulty.
Choice of Shape for Use With the Glaze
Crystalline glazes have traditionally been applied to simplified, smoothly curving symmetrical forms. Platters and bottles or other convex forms have been considered to be well suited to crystalline glazes as they provide large areas of unbroken surface on which to display crystals.
Zinc-silicate crystals may be quite large and, consequently, may require large scale pieces to bear their visual weight. Such crystals are used most successfully on large, unbroken, curved areas.
The moly crystal is entirely different in visual character from the zinc-silicate crystal. Moly crystals rarely exceed one inch in diameter and are always delicate and visually elusive. They are often not apparent on the surface of a glaze until it is closely observed. These small, precious crystals, then, require special consideration in the designing of the ware which will carry them.
The moly crystals produced in glaze CFM will not grow on a purely horizontal surface such as a platter. It is possible that such a flat surface does not permit enough movement in the glaze to generate crystals. The plate form, then, frequently used with zinc crystalline glazes, cannot be used with glaze CFM.

Bowl forms are rarely used with zinc-silicate crystal glazes as the thick glaze pool which inevitably collects in the bottom of the bowl usually produces thick, matted crystals which are rough to the touch. The moly crystal glaze, however, seems particularly well adapted to the bowl form. When fired correctly, the glaze pool in the bottom of the bowl will be texturally smooth and visually fascinating, exhibiting a complicated, interwoven pattern of the crystals.
Like zinc-silicate crystals, the moly crystal adapts well to the bottle or vase form. The small moly crystals do not, however, demand the large scale of objects which the larger zinc crystals require. A large form may in fact dwarf the small moly formations. A small piece which may be held in the hand, turned to the light and closely examined, may provide the best view of the varied shape and elusive coloration of these crystals. Although the placement of crystals on a piece is unpredictable, moly crystals tend to pool and collect attractively on areas of small detail which contain a horizontal plane. Regardless of the scale or specific shape of the piece bearing moly crystals, a form which is appropriate to the glaze must emphasize the delicate and precious nature of the crystal.
The irridescent qualities of the moly crystal may be enhanced and intensified by exposing the fired glaze to strong acid fumes in a process known as fuming.

Irridescence is an optical effect produced by the interference of light rays diffracted and reflected from two
glazes which are separated by a few wavelengths in distance. This effect may be achieved by subjecting a fired glaze to the acid fumes of a number of substances, such as stannous chloride, ferrous chloride or barium nitrate, in a closed chamber at approximately 1290 F (700 C) The temperature of the chamber must be carefully controlled, as a temperature much above or below 12 90 may cause an unattractive scum to form on the surface of the glaze.
Only small amounts of the fuming agents are required. Two teaspoonfuls proved sufficient when introduced through a peephole and deposited directly on the kiln shelf of a two cubic foot kiln. A small draft should be maintained to provide circulation of the vapors throughout the kiln. Adequate ventilation of the area around the kiln must be provided as the fumes are extremely irritating, and precautions should be taken to avoid their inhalation.
C. W. Parmelee, Ceramic Glazes, 541.

The moly crystal described in this investigation has not been previously quantitatively or qualitatively characterized. Characterization of a material, as described by the Materials Advisory Board of the National Research Council "...describes those features of the composition and structure (including defects) of a material that are significant for a particular preparation, study of properties, or use, and suffice for the reproduction of the material.""'' Characterization of the moly crystal, then, must include an analysis of its composition and structure sufficient to allow its reproduction.
Research techniques currently available provide means for gathering information about the chemical, structural, microstructural and surface characterization of glasses, including ceramic glazes. The analytical methods used in conjunction with this investigation were X-Ray Diffraction (XRD), Infrared Reflected Spectroscopy (IRRS), Electron Microprobe (EMP), Scanning Electron Microscopy (SEM), and Aujer Electron Spectroscopy (AES). While each analytic technique provided limited and specific information, the aggregate of data produced an identification of the crystal
L. L. Hench, "Characterization of Glass", Characterization of Materials in Research, Ceramics and Polymers, Syracuse,
N.Y.> 1975, 1.

phase of glaze CFM, as well as information on the nucleation of the crystal and an understanding of the compositional and thermal conditions necessary for crystal growth.
Experimental Instruments X-Ray Diffraction (XRD)
X-Ray Diffraction (XRD) provides the primary means for characterization of atomic or ionic structure of crystalline materials. The atoms of a crystalline solid are fixed in a precise, periodically repeating pattern, while the atoms in a glassy lattice are randomly distributed. In XRD analysis, an x-ray beam is focused on a rotating sample. The x-rays will diffract from the interplanar spacings in an organized solid such as a crystal, producing recorded peaks. At the energy level used in this procedure no diffraction, and therefore no significant peaks, will result from aiming the beam on the nearly featureless structure of glass. Therefore, in a sample containing both glassy matrix and crystal, a crystalline area may be easily discerned from a glassy matrix in the analysis of results.
Rotation of the sample is necessary in order to increase the chances of the crystal giving rise to diffraction. Grinding the crystal to a fine powder further increases the variety of orientations of the crystal, thereby increasing the probability of diffraction and the production of definitive peaks. These peaks may be indexed and compared to those produced by known crystals, so as to identify the atomic structure of the sample material.

Infrared Reflection Spectroscopy (IRRS)
IRRS is a quantitative tool used for characterizing surface structure and composition of glasses. When excited, the vibration or rotation of various molecular species will generate particular absorption bands in the infrared light spectrum. When a particular molecular species is contained on a surface, as with a crystal contained in a glassy matrix, the surface and molecular species will each produce characteristic infrared spectrum bands. The separate bands provide information about the composition of the molecular species in question and about the manner in which the surface environment may modify that species.
Electron Microprobe (EMP)
The Electron Microprobe can provide an analysis of the microstructure of glass. The prepared sample, contained in a vacuum chamber, is excited by an electron beam which produces the emission of an x-ray from the atomic structure of the sample material. The x-ray emission, which is specific to each element, may then be analyzed. By this method, the technique provides an identification of the elements in the sample and a quantitative estimate of the concentration of each element in the sample.

Although EMP provides an analysis of approximately 2 3
2 to lOym of depth, this is essentially only a surface analysis. Any preparation of the sample surface which might alter its chemical composition may result in ambiguous or inaccurate results. In addition, EMP will tolerate only a certain degree of surface roughness in the sample which may prove to be a problem in the analysis of a crys-alline glaze.
Scanning Electron Microscope (SEM)
The Scanning Electron Microscope, or SEM, is a tool widely used for microstructural analysis. In the SEM, a focused electron beam scans the surface of a prepared sample contained in a vacuum chamber. The primary electron beam strikes the surface of the sample which emits secondary electrons. The morphology of the sample surface determines the number of secondary electrons emitted. The resultant emission current provides a visual display which can be used to provide a qualitative visual microstructural analysis. The SEM is capable of a large depth-of-field focus and can provide magnification of lOx to 50,000x. When used in conjunction with emitted x-ray analysis, SEM can provide a qualitative compositional analysis of the sample surface. These data can be made semiquantitative
lum = 10 m
L. L. Hench and R. W. Gould, ed., Characterization of Ceramics, New York, 1971, 506-507.

by comparison with standards of known composition. Aujer Electron Spectroscopy (AES)
Aujer Electron Spectroscopy, or AES, is a qualitative tool which provides a chemical analysis of the surface composition of a sample. The data provided by AES often supplements that obtained via the Electron Microprobe.
In Aujer Electron Microscopy, the surface of a sample material, contained in a vacuum chamber, is excited by bombardment by an electron beam. The electrons act to ionize one of the electron levels of the atoms in the material. As the atomic system tries to reestablish equil brium, a series of electron shifts from shell to shell causes an electron known as the Aujer electron to be released and ejected from the sample. The released Aujer electron is characteristic of the atom which ejects it, and it may be identified by the energy levels at which recorded peaks occur.
All elements but hydrogen and helium are identifiable
with AES, and the technique is highly sensitive. The
Aujer electron can only migrate to the surface and escape
from a depth of a few atomic layers, however. This limits
the examination of the sample surface to a depth of only a
few monolayers, or about 50A.
1A = 10 cm

.Ion bombardment with an element of high mass number, such as argon, can provide information on deeper layers of the sample. As the sample chamber is filled with argon gas and excited by an electron beam for predetermined periods of time, the atomic layers of the sample surface are etched away, and analysis of the lower layers may be carried out by obtaining an AES analysis of the argon etched surface.
Experimental Procedures
Several glaze samples were prepared for XRD analysis. Sample 1 was a small glazed tile, approximately 2 cm by 2 cm, containing areas of crystal and glaze. Samples 2 and 3 were powder preparation slides. Sample 2 contained ground material from a glazed tile which contained both glass and crystal. Sample 3 contained ground materials from a test piece which had, because of improper firing conditions, produced a profusion of nucleated but underdeveloped crystals. Each of the three samples was mounted in a sample holder and rotated between 2 and 70 while being bombarded with x-rays. The diffraction peaks and corresponding 2--angles were recorded and indexed. IRRS
An "as fired" sample tile containing both crystal and glaze was examined with IRRS. One spectrum was obtained from the glassy area and another from an area containing a large crystal, The peaks obtained were compared with

the results of previous scientific investigations. EMP
A flat sample bearing both crystals and glaze was carbon-coated to avoid electrical charging effects during analysis. A lOOym electron beam was scanned across the surface at a rate of 100pm per minute to prevent diffusion of mobile species. Counts for Si, Ca, Zn, Ti, K, Al, and Mo were taken at the equidistant locations from approximately the center of a crystal across the crystal-glass interface to the surrounding glassy phase of the glaze over a total distance of approximately 5mm. The number of counts for each element were plotted vs. the distance across the sample surface giving a compositional profile of the crystal-glaze interface. SEM
The carbon coated sample used for EMP analysis was mounted onto an appropriate sample holder and placed into the SEM vacuum chamber. The surface was examined visually from 20x to 5,000x for morphological detail. Energy dispersive x-ray analysis was used to examine several regions in the crystal phase. The elements producing each peak in the EXDA spectra were determined, and the relative heights of the peaks were recorded. AES
A small sample of glaze was cut from a fired tile and prepared for mounting in the vacuum chamber. Analysis of
the near surface (ca. 50A) of the glass and crystal was

produced. Argon ion milling was used to remove layers of the surface of the sample to determine the composition
o o
at depths of approximately 1,500A and 2,500A. The AES differential spectra peaks were indexed, and peak-to-peak element to oxygen ratios were calculated.
Experimental Results
Sample 1 produced too few peaks for definitve identification of the crystal structure. These equivocal results probably resulted from the large ratio of glass to crystal contained on the sample and from the fixed orientation of crystals in the sample tile.
Sample 2 produced peaks of very low intensity which caused difficulty in determining which peaks were significant for indexing purposes. It is likely that a large amount of glass contained in the sample with the crystals was responsible for the problem.
Sample 3 produced the most easily interpreted results. The diffraction pattern produced by sample 3 is shown in Figure 5. Table 3 lists the individual d-values and comparative intensities for the most prominent peaks of the pattern. In Table 4 the sample d-values and intensities for sample 3 are compared with d-values and intensities for calcium molybdate (CaMoO^) and other possible crystal structures. As the table shows, there is very close agreement in the unknown sample and CaMoO^. Such close agreement in the indexing of the unknown crystal by this

Figure 5. X-Ray Diffraction Pattern Produced by Glaze Sample


Table 3. 2-e- Angles, d-Values and Intensities for Glaze Sample 3.
2-e-_d-Values_ Intensity
18.7 4.75 Strong
21.85 4.07 Weak
26.75 3.33 Weak
28.85 3.09 Very Strong
31.32 2.86 Very Weak
34.42 2.61 Medium
47.25 1.92 Medium
49.25 1.85 Weak
54.28 1.69 Very Weak
58,1 1,59 Weak

Table 4. XRD d-Values and Intensities from Glaze Sample 3 Compared with d-Values and Intensities of Calcium Molybdate (MoO.) and Alternative Crystals.
Calcium Molybdate
Sample 3__CaMo04a_Na2Np (M0O4 ) 3_CaWo4
d-Value Intensity d-Value I/IO d-Value Intensity d-Value I/Io
4.75 Strong 4.76 28 4 .72 Not Available 4.76 53
4.07 Weak
3.33 Weak
3.09 Very Strong 3.10 100 3 .10 3.10 100
2.86 Very Weak 2.86 13
2.61 Medium 2.61 16 2.61 2. 62 23
1.92 Medium 1.929 30 1.93 1.93 28
1.85 Weak 1.848 13 1.84
1. 69 Very Weak 1.694 14 1.69 1.69 16
1. 59 Weak 1.588 22 1.59 1.59 10
a card #7- 212.
b d-Values and I/Io from American Society for Testing Materials Indes Card #7-210.

technique essentially confirms the identification of the
unknown moly crystal as calcium molybdate.
Figure 6 shows the IRRS spectrum obtained from the glaze sample. The spectrum obtained from the glassy area is typical of silicate glasses. A broad peak centered at 1,050cm is due to the stretching of Si-0 bonds of the silicontetrahedra. Another peak at 650cm ^ is due to the rocking of the same Si-0 bonds in the silicon tetrahedra.
The second spectrum was obtained from a portion of
the glaze containing a large crystal. As in the first
spectrum, peaks at 1,050cm ^ and 650cm ^ were observed.
These are due to the glassy phase present around the crystal.
Two other peaks at 900cm ^ and 400cm ^ are also observed.
Infrared absorption studies have shown that stretching of
Mo-0 bonds in the MoO^ tetrahedra produce a stong peak
-1 -1 5
at 840 810cm and 340 320cm .
The results of the Electron Microprobe study are shown in Figure 7, which plots the number of counts for each element at the various points from the center of the crystal to the surrounding glassy phase. The number of counts is proportioned to the concentration of each element. The location of the crystal-glass interface is indicated in
C. N. R. Rao, Chemical Applications of Infrared Spectroscopy New York, 1963, 355.

Figure 6. IRRS Spectrum Obtained from Glaze Sample

Figure 7. EMP Microstructural Analysis of Glaze.

600 A
500 A
00 h-
o o
400 -
300 A
ioo A
Si x 10
-O Q li
1-1-1-1-T ? 9 -r
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
Distance from Center of Crystal (mm)

the figure. As the graph shows, most of the calcium and molybdenum are concentrated in the crystal, while titanium and zinc occur in larger quantities in the glassy phase. Potassium, aluminum and silicon show nearly equal counts in the glass and the crystal, SEM
While visual examination of the sample provided no significant data other than to locate the crystal and crystal glass interface, x-ray studies performed with SEM provided additional compositional analysis of the sample. Table 5 shows the quantitative results obtained with this method. The data indicate that larger concentrations of silicon, potassium, titanium, barium and copper^ are present in the glassy portion of the glaze. Zinc appears to be equally distributed across the sample. Calcium and molybdenum are present in larger concentrations in the crystal. The small iron peak is the result of electron scattering in the tube and does not represent the average composition of the glaze. AES
Table 6 lists peak-to-peak element to oxygen ratios for the crystal and glassy phases of the glaze at indicated depths obtained by AES. The results indicate that silicon and aluminum are present in large amounts in the glassy phase. Potassium, zinc, barium, calcium and titanium are
Copper carbonate was used as a colorant in the glaze.

Table 5. SEM Emitted X-Ray Compositional Analysis of Glaze Sample
_Peak Size Ranking_
1 2 3
Outside of Center of Peak_Peak Size_Element_Glaze_Crystal_Crystal
1.75-1.81 Very large Si K oc 1.74 1 2 > 3
2.34-2.35 Small Mo L cc 2.29 1 < 2 < 3
3.33-3.34 Medium K K cc 3.31 1 > 2 > 3
3.69-3.72 Large Ca K oc 3.69 1 < 2 < 3
4.02-4.03 Small Ca K 6 4.01 1 2 s 3
4.51-4.52 Medium Ti Ti K K oc B 4.51 4.93 1 2 > 3
4.92-4.94 Small Ba L 6 4.83 1 2 > 3
6.37-6.41 Small Fe K oc 4.40 1 2 3
7.98-8.06 Small Cu K cc 8 .03 1 > 2 3
8.62-8.88 Small Zn K cc 8.64 1 2 3

Table 6. AES Peak-to-Peak Element to Oxygen Ratios.
Glazeeat Crystal Crystal at __1500A_Surface_1500A
Si:0 .074
K:0 .009 .081
B:0 Trace
Ti:0 .011 .023
Ba:0 .008 .005 A1:0 .034
Ca:0 .034 .731 .642
Mo:0 .185 .219

also present in the glass. The crystal surface shows a predominance of calcium and molybdenum. Traces of barium and titanium are also indicated. Milling a very small
distance, to a depth of 1,500A, indicated that calcium and molybdenum are the only significant constituents of the crystal. The presence of titanium and barium on the surface was probably due to contamination arising from handling, sample preparation, or diffusion onto the surface of the crystal during firing.
Discussion of Results The data collected suggest that the moly crystal is identified as calcium molybdate, CaMoO^. The x-ray diffraction pattern produced by the crystal phase strongly supports
this identification. As Table 4 indicates, eight of the
peaks produced by the sample match the ASTM index of
values for calcium molybdate. The alternative crystal
structures listed in Table 4 show fewer matching peaks.
Knowledge of the glaze composition eliminates most of these
alternative crystal structures even when the possibility of
element substitution within crystals is considered.
Data gathered from IRRS, EMP, SEM and AES confirm
identification of the crystal phase as calcium molybdate.
As shown in Figure 6, the IRRS spectrum identified Mo04 as a major structural component of the crystal phase, but not of the glassy phase. EMP, SEM and AES indicate the
American Society for Testing Materials.

presence of both calcium and molybdenum in the crystal phase. Quantitative data provided by these three instruments indicate that while some calcium exists in the glassy phase, the greater quantity of calcium, and nearly all the molybdenum in the glaze is contained in the crystal phase. This data is recorded in Figure 7 and Tables 5 and 6.
While AES suggested that calcium. ;and molybdenum were the only significant elements appearing in the crystal phase, as shown in Table 6, the Electron Microprobe study indicated about equal amounts of aluminum and silicon over the ten data points in the crystal and glassy phase as seen in Figure 7. These seemingly conflicting pieces of data may be explained by the fact that the AES analyzed the sample
to a total depth of 2,500A within the crystal, while EMP probed the samples to a depth of approximately 2-10 urn. Data obtained by EMP analysis represents an average of the material to this depth under the beam. If the crystal phase was less than 1 vim thick, the EMP would provide compositional data on both the crystal and the glass below it. The extremely thin growth habit of the moly crystal was exhibited when a sample was immersed in hydrofluoric acid for three seconds, resulting in the total dissolution of the crystal from the glaze surface. Data generated by an. AES probe to a depth
of 1,500A provides more accurate information on the constitution of the crystal phase exclusively.
SEM, AES and EMP all indicated that titanium was not concentrated in the crystal, but remained in the glassy phase

of the glaze. Experimentation with the moly crystalline glaze had revealed, however, that the pressence of titanium was necessary for the formation of crystals. The phase equilibria diagram shown in Figure 8 explains the necessity of titanium in the process of crystal formation. In a K^TiO^ K^MoO^ glaze system, two immiscible liquids are formed above 917C. Upon soaking at a temperature above 917C during the cooling cycle, molybdenum-rich areas are produced which, after nucleation of the crystal, grow to form macro-crystals. Titanium is therefore necessary in the glaze melt in order to produce the phase separation which ultimately results in the formation of the moly crystal.

Figure 8. Phase Equilibria Diagram for ^TiC^-I^WC^-I^MoO Na2Ti03-Na2Wo4, Na2Ti03-Na2Mo04 Gldze Systems.

7 4
K2Mo04 926
Systems: 1) K2Ti03-K2W04; 2) K2Ti02-K2Mo04 ;
3) Na2Ti03-Na2W04; 4) Na2Ti02-Na2Mo04
Source: M. L. Sholokhovich and G. V. Barkova,Zhur. Obshchei Khim., 25, 1258, 1956.--

The moly crystal is a unique crystal formation. By learning to control its growth and development in a glassy matrix, it becomes possible to produce a new species of crystalline glaze. The compositional elements and firing conditions required for the production of the moly crystal in a high temperature stoneware glaze have been empirically determined in this investigation. Characterization of the crystal has provided a better understanding of these requirements and has confirmed much of the empirically generated data.
Prior to this investigation, the association of molybdenum with this particular crystal formation had been presumptive. The results of this study, however, specifically identify the crystal phase as calcium molybdate, CaMoO^, and thus document that calcium is an essential constituent of the glaze. The necessity and function of titanium in the glaze melt has also been explained. This information may provide the basis for further development of moly crystalline glazes.
The glaze developed in this investigation requires careful preparation and treatment in order to assure the production of well-formed, macro-crystals in a glaze surface which is free of defects. Nevertheless, the glaze is

practical for use by the ceramic artist who is willing to take the time which the preparation of a crystalline glaze demands.


Firing Schedule #1. Globar Silicon Carbide Element Kiln.
Time Setting Temperature3
5:00 p.m. 2
7:00 p.m. 4
9:00 p.m. 6
8:00 a.m. 8 cone 5 down
10:00 a.m. off cone 11 down
10:15 a.m. 4 2300F
11:00 a.m. 4 1825F
3:00 p.m. 4 1600F
3:30 p.m. off 1575F
Temperature determined by pyrometer readings or by observation of pyrometric cones.

7 9
Firing Schedule #2. Globar Silicon Carbide Element Kiln.
Time_ Setting_Temperature
3:00 p.m. 2
5:00 p.m. 4
9:00 p.m. 6
6:00 a.m. 8 1700F
8:00 a.m. 8 1975F
8:30 a.m. 8 cone 5 down, cone 7
8:57 a.m. 7 cone 9 halfway down, 2
9:30 a.m. 4 2300F
10:00 a.m. 4 2100F
10:40 a.m. 4 2000F
11:15 a.m. 4 1950F
11:30 a.m. 4 1900F
11:50 a.m. 4 1850F
12:30 p.m. 4 1750F
1:10 p.m. 4 1750F
2:15 p.m. 4 1675F
3:15 p.m. 4 1620F
4:20 p.m. Off 1550F
Temperature determined by pyrometer readings or by observation of pyrometric cones.

Firing Schedule #3. Globar Silicon Carbide Element Kiln.
Time Setting Temperature
7:00 p.m. 2
9:30 p.m. 4
11:00 p.m. 6
8:00 a.m. 7 1700F
10:00 a.m. 7 1900F
12:00 p.m. 7 cone 2 down
1:15 p.m. 7 cone 5 down
2:15 p.m. 7 cone 8 down
3:20 p.m. 4 cone 9 down
3:40 p.m. 4 2200F
4:10 p.m. 4 2075F
4:40 p.m. 4 2060F
5:10 p.m. 4 2020F
6:10 p.m. 4 1960F
8:10 p.m. 4 1750F
10:15 p.m. off 1650F
Temperature determined by pyrometer reading or by observation of pyrometric cones.

Firing Schedule #4. Olympic Kanthal -Wire Element Kiln.
Time Setting Temperature3
9:00 a.m. low
11:00 a.m. 1
12:30 p.m. 3.5 1200F
2:00 p.m. 3.5 2000F
2:40 p.m. off 2200 F, cone 8 dov
3:00 p.m. 2.3 2000F
3:40 p.m. 2.3 2000F
4:40 p.m. 1.5 1900F
5:30 p.m. 1 1850F
6:00 p.m. off 1800F
Temperature determined by pyrometer readings or by observation of pyrometric cones.

,8 2
Firing Schedule #5. Olympic Kanthal- -Wire Element Kiln.
Time Setting Temperature3
7:00 a.m. low
7:45 a.m. 2 600 F
9:00 a.m. 3 1000F
9:30 a.m. 4 1200F
10:15 a.m. 5 1700F
11:00 a.m. 5 2000F
11:45 a.m. off 2240F, cone 9 do\
11:50 a.m. 2.5 2100F
12:30 p.m. 3 2040F
1:00 p.m. 2.75 2120F
1:30 p.m. 2.75 2075F
2:00 p.m. 3.5 2075F
2:30 p.m. 3.5 2100F
2:45 p.m. 3.5 2150F
3:15 p.m. 3.5 2150F
4:00 p.m. off 2150F
Temperature determined by pyrometer readings or by observa tion of pyrometric cones.

Firing Schedule #6. Olympic Kanthal-Wire Element Kiln.
7:00 a.m. low
8:30 a.m. 2 700F
11:45 a.m. 4 1400F
2:20 p.m. off 2260F, cone 9 down
2:50 p.m. 2.5 2000F
4:10 p.m. 3 1950F
7:00 p.m. off 2000F
Temperature determined by pyrometer readings or by observa-
tion of pyrometric cones,

Firing Schedule #7. Globar Silicon Carbide Element Kiln.
Time Setting Temperature
3:30 p.m. 2
7:00 p.m. 4
10:00 p.m. 6
8:00 a.m. 8
8:20 a.m. 8 cone 2 down
8:45 a.m. 8 cone 5 down
9:00 a.m. 7 cone 7 down
9:45 a.m. 8 2050F
10:10 a.m. 8 2150F
10:55 a.m. off 2190F, cone 9
11:15 a.m. 5 1995F
12:30 p.m. 5 1920F
1:10 p.m. 6 1900F
2:10 p.m. 7 1900F
2:30 p.m. 8 1850F
2:40 p.m. 7 1950F
4;15 p.m. off 1900F
Temperature determined by pyrometer readings or by observa tion of pyrometric cones.

Firing Schedule #8. L+L Kanthal-Wire Element Kiln.
Time Setting Temperature3
7:00 p.m. low
10:00 p.m. medium
9:45 a.m. high
10:45 a.m. high cone 7 down
12:00 p.m. off 2100F, cone 9 down
12:20 p.m. medium 1850F
2:00 p.m. high 1750F
2;30 p.m. medium 1850F
4:00 p.m. off 1840F
a Temperature determined by pyrometer readings or by observation of pyrometric cones.

Altrom, Myron J., The Development of Molydic Oxide Crystals
and Their Use in Decoration on Porcelain, Master's Thesis, San Jose State College, 1960.
"Annual Raw Material Processing Handbook", Ceramic Industry, Chicago, Illinois, 1, 1971.
Hench, L. L. "Characterization of Glass", Characterization of Materials in Research, Ceramics and Polymers, Syracuse, New York, 1975.
Hench, L. L. and Gould, R. W., ed., Characterization of Ceramics, New York, New York, 1971.
Hennessey, Linda, B., Factors Affecting the Growth of Zinc
Silicate Crystals in a Glaze, Master's Thesis, University of Puget Sound, 1974.
Kahn, Annelies, Molybdenum as a Crystal Producing Agent in a Stoneware Glaze, Master's Thesis, Texas Women's University, 1970.
Kraner, H. M., "Colors in a Zinc-Silicate Glaze", Journal of the American Ceramic Society, Columbus, Ohio, 7, 1924.
Lawrence, W. G., Ceramic Science for the Potter, New York, New York, 1972.
Nelson, Glenn, Ceramics: A Potter's Handbook, New York, New York, 1971.
Norton, F. H., "The Control of Crystalline Glazes", Journal of the American Ceramic Society, Columbus, Ohio, 20, 1937.
Parmelee, C. W., Ceramic Glazes, Boston, Massachusetts, 1973.
Rao, C. N. R., Chemical Application of Infrared Spectroscopy, New York, New York, 1963.
Sanders, H. H., Glazes for Special Effects, New York, New York, 1974.
Snair, D., "Making and Firing Crystalline Glazes", Ceramics Monthly, Columbus, Ohio, December, 1975.

BIOGRAPHICAL SKETCH Dorothy Rita Bassett was born on April 9, 1949, in White Plains, New York. She graduated from Scarsdale High School, Scarsdale, New York in 1967. In 1971 she received a Bachelors degree from the University of Pennsylvania, Philadelphia, Pennsylvania where she studied history and art history. She earned a Master of Arts in Teaching degree from the University of Virginia, Charlottesville, Virginia in 1972.
From 1972 to 1976, when she entered graduate school at the University of Florida, Ms. Bassett was an Instructor of Art at Piedmont Virginia Community College, Charlottesville, Virginia. As a Graduate Assistant at the University of Florida, she taught ceramics while working towards the degree of Master of Fine Arts.

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a thesis for the degree of Master of Fine Arts.
I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a thesis for the degree of Master of Fine Arts.
Eugene E. Grissom Professor and Chairman, Department/ of Art
I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a thesis for the degree of Master of Fine Arts.
^ / ........-:. '^XZ
; Larry y. Herfen'
Professor and Head of Ceramics Division, Department of Materials Science and Engineering
This thesis was submitted to the Graduate Faculty of the College of Fine Arts and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Master of Fine Arts.
June 1978
Dean, Graduate School

sm2 A
78. *

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd